Challenges in the future of cancer screening.

The purpose of cancer screening is to reduce mortality, and ideally incidence, from the cancer screened for. Until recently, cancer screening has been offered to all persons in pre-defined sex- and age-groups. The exception is lung screening which is targeted to high-risk individuals. Evidence for effect of screening on cancer-specific mortality is available for cancer of the cervix, breast, colorectal, lung, and prostate, and on cancer-specific incidence for cervix and colorectal cancer. With more effective molecular and computational tools, the current paradigm of cancer screening will change. Manual reading of images and pathology in mammography, cytology, colposcopy, colonoscopy, and CT scan will be complemented or replaced by AI-interpretation. Risk stratification will diversify screening intensity, for instance in breast screening where modelling and randomized controlled trials are underway. Blood-based screening tests might allow for simultaneous early detection of several types of cancer. Furthermore, prediction models based on life trajectories in health and other data will enhance risk stratification, potentially dividing the population into groups with no need of screening, with need of simple or advanced screening, with need of surveillance or even diagnostics. In public health care systems, these developments must be closely monitored. Before replacing an existing program, evidence for non-inferiority in reducing cancer-specific mortality should be ensured. Benefits must outweigh harms, and citizens should have confidence in new screening schemes. With the pressure on health care resources, screening should continue in organized and monitored programs. Taking these conditions into account, the new screening tools will potentially enable improved cancer control.